Education Principles
The immutable truths. Institutions change. Technology evolves. These don't.
The Five Principles
| # | Principle | Why Immutable | Implication |
|---|---|---|---|
| 1 | Agency compounds | Pattern skills stack exponentially (n² effect) | Early agency → lifetime advantage |
| 2 | Shipped work proves | Output can't be faked like credentials | Portfolios replace papers |
| 3 | Coaching unlocks | Belief precedes capability | Great coaches > great content |
| 4 | Networks multiply | Talent discovers talent (Metcalfe) | Masterminds > classrooms |
| 5 | Games train patterns | Tightest feedback loops exist in play | Games as serious pedagogy |
1. Agency Compounds
The earlier you develop agency, the longer it has to compound. Like financial capital, pattern capital grows exponentially.
The math: Pattern skills don't add linearly—they multiply. Recognition × Utilization × Creation = Agency. Each skill amplifies the others.
The implication: Four years building agency beats four years accumulating credentials. The credential holder starts at zero post-graduation. The agency builder has compounding momentum.
What agency looks like:
| Pattern Skill | What It Is | Education Builds It Through |
|---|---|---|
| Recognition | See patterns before others | Sports (reading play), market games |
| Utilization | Act on what you see | Ship projects, real consequences |
| Creation | Develop new patterns | Sandbox exploration, game design |
The gap: Traditional education trains recognition (tests, analysis) but neglects utilization and creation. Agency accelerators train all three.
2. Shipped Work Proves
Credentials verify input (time spent, tests passed). Portfolios verify output (value created, problems solved).
Why it's immutable: AI can generate credentials. AI can pass tests. AI cannot ship work that creates value for others. Shipped work requires judgment, coordination, and accountability—human skills.
Traditional vs Agency:
| Dimension | Credential | Shipped Work |
|---|---|---|
| Verifiable by | Institution | Anyone |
| Fakeable | Increasingly | No—results are visible |
| Transferable | Within system | Universal |
| Compounding | Depreciates | Compounds (portfolio effect) |
The implication: Every project shipped is permanent evidence. Every course completed is temporary signal. Over a career, portfolios win.
The formula: Proof of Work = (Shipped Project) × (Visible Impact) × (Verifiable Attribution)
3. Coaching Unlocks
Content is commodity. Coaching is transformation. The difference is belief.
The insight: A great coach doesn't teach technique—they sell belief in untapped potential. The student who believes they can already learned the hardest part.
Zone of Proximal Development (ZPD)
Learning happens in the gap between what you can do alone and what you can do with help. Not in what you already know. Not in what's too far beyond. In the zone between—where you need a guide.
CAN'T DO (even with help)
│
│ ┌─────────────────────────────┐
│ │ ZPD - CAN DO WITH GUIDANCE │ ← Learning happens HERE
│ │ ┌─────────────────────┐ │
│ │ │ CAN DO ALONE │ │
│ │ │ (current capability)│ │
│ │ └─────────────────────┘ │
│ └─────────────────────────────┘
│
More Knowledgeable Other (MKO)
The person just ahead of you—not an expert miles away, but someone who recently crossed the gap you're facing. The best teacher isn't the most knowledgeable. It's the one who can bridge YOUR specific gap.
The role flips: A father learns to read the game from his son in Fortnite. The son learns to read life from his father. Everyone is simultaneously teacher AND student—just in different domains. "Teacher" isn't a role. It's a contextual relationship that flows wherever the ZPD exists.
Why coaching beats content:
| Content | Coaching |
|---|---|
| Same for everyone | Personalized to individual |
| Answers questions | Asks better questions |
| Transmits knowledge | Transfers belief |
| Scales infinitely (AI) | Scales through network effects |
The pattern: Knowledge → Skill → Identity. Content can transfer knowledge. Only coaching can shift identity.
The coaching moment:
Coach recognizes pattern in student
↓
Names the pattern they can't yet see
↓
Student sees themselves differently
↓
New identity → new capability
AI's role: AI delivers content at infinite scale. This makes human coaching MORE valuable, not less. The gap to fill is belief, not information.
4. Networks Multiply
Education has always been two products bundled together: content and network. AI unbundles content. Network becomes the primary value.
The math (Metcalfe's Law): Network value = n². Double the quality peers, quadruple the opportunity.
Traditional model:
Institution gatekeeps → Network = classmates → Value = who you met
Agency accelerator model:
Shipped work → Portfolio attracts → Network = collaborators → Value = who knows your work
Why networks compound:
| Network Type | How It Forms | Value Capture |
|---|---|---|
| Alumni | Shared institution | Weak—shared time, not shared work |
| Collaborator | Shared projects | Strong—mutual skin in the game |
| Mastermind | Shared problems | Strongest—ongoing reciprocity |
The implication: Masterminds where people solve problems together beat classrooms where people listen together. The unit of education shifts from "cohort" to "crew."
5. Games Train Patterns
Prediction is fundamental. The ability to read what's about to happen—then act on it—is the core skill of agency. Games are how humans have always trained this.
Why games:
| Property | Why It Matters |
|---|---|
| Tight feedback loops | Learn faster when consequence follows action |
| Safe failure | Mistakes cost points, not livelihood |
| Pattern density | More pattern exposure per hour than real life |
| Flow state | Optimal challenge/skill ratio |
Pattern skill development through games:
| Pattern Skill | Game Type That Trains It | Why It Works |
|---|---|---|
| Recognition | Sports (reading play), strategy | Immediate feedback on what you saw vs what happened |
| Utilization | Team sports, coordination games | Use patterns to achieve shared objectives |
| Creation | Sandbox games, game design | Build systems, test novel patterns |
Sports as prediction training:
The rugby player reading the play. The basketball point guard seeing the court. The chess player anticipating moves. These aren't separate from "academic" skills—they're the foundation.
Reading play = pattern recognition + prediction + action under uncertainty.
This is the same skill as:
- Reading a market
- Reading a room
- Reading a codebase
- Reading what a customer actually wants
The implication: Games before homework improves retention. Regular gamers anticipate consequences better than non-gamers. Gaming teaches collaboration with systems—increasingly essential as AI becomes teammate.
The Test
Before any educational investment or design:
| Question | Yes = Proceed | No = Reconsider |
|---|---|---|
| Does this build agency? | Pattern skills compound | Knowledge only |
| Does this produce shipped work? | Portfolio grows | Credentials only |
| Does this include coaching? | Belief transfers | Content only |
| Does this build network? | Connections compound | Isolated learning |
| Does this use games? | Tight feedback loops | Lecture-based |
Minimum: Yes to 3 of 5.
The Knowledge Types
Education must develop all three forms of knowledge:
| Type | What It Is | How It's Learned | AI Role |
|---|---|---|---|
| Embodied | Knowledge in tools/materials | Using tools | AI augments |
| Codified | Knowledge in symbols | Reading, instruction | AI delivers |
| Knowhow | Knowledge in people | Practice, coaching | AI can't replace |
Traditional education over-indexes on codified knowledge—the exact type AI delivers cheaply.
Agency accelerators develop knowhow—the type AI cannot replace.
Principles → Performance
These principles determine what to measure:
| Principle | Performance Metric |
|---|---|
| Agency compounds | Pattern skill scores, agency assessments |
| Shipped work proves | Ship rate, portfolio depth, verified projects |
| Coaching unlocks | Coach match rate, belief shift metrics |
| Networks multiply | Network size, collaboration density |
| Games train patterns | Pattern recognition scores, prediction accuracy |
See Performance for the full metrics framework.
Context
- Education Overview — The transformation thesis
- Agency — Why agency matters
- Play — The heroes journey framework
- Games — Pattern training ground
- 2026 Review — The education split prediction
- Prediction Process — How to build prediction skill